Latest news with #Processing
Euractiv
19 hours ago
- Business
- Euractiv
Energy's quantum leap will power Europe's supercomputers
As big data centres and artificial intelligence generate continued transformations in all areas of society, there has been increasing attention paid to how much energy these supercomputing activities use. But what is less known is that these major computing operations are also changing how energy is generated. From optimising oil and gas exploration to accelerating breakthroughs in renewables and fusion energy, advanced computing is transforming how the energy sector operates. High-performance supercomputers capable of performing immense numbers of calculations per second are finding new applications in the sector. Traditionally, such computing power has been used to analyse massive seismic datasets, helping geologists locate new oil and gas reserves deep beneath the earth's surface. But today, supercomputers are tackling far more - simulating new battery chemistries, designing more efficient solar panels, and modelling carbon capture and storage projects that could trap millions of tonnes of CO₂ underground. Energy pioneers Much of the pioneering work in this area has been going on in Europe. For instance, Italian energy company Eni has, over the past decade, been making rapid advances on its supercomputers. For years, starting in the 1970s, Eni had used supercomputing on various fronts in its traditional businesses, from optimising the operation of industrial plants to simulating the behaviour of plasma in magnetic confinement fusion. But since 2013, Eni has equipped itself with successive generations of supercomputers housed in Eni's Green Data Centre in Ferrera Erbognone, near Pavia. 'On supercomputing, it's been a very long journey at Eni, continuing today, because we started many years ago,' says Dario Pagani, Head of Digital and Information Technology at Eni, and CEO of the company's joint venture Eniquantic. 'More than ten years ago, Eni was among the first companies to believe in the so-called 'hybrid' supercomputing architectures, those combining CPUs (Central Processing Units) with GPUs (Graphics Processing Units). GPUs, initially conceived only for massive parallel computing in gaming consoles, are now at the core of the most advanced supercomputers in the world. Eni has therefore had a pioneering vision on this, and we are now reaping the fruits in terms of competitive advantage and innovation possibilities.' Eni isn't alone in Europe in its interest in the supercomputing space. TotalEnergies in France has its Pangea supercomputers, which combine on-site high-performance computing with cloud. in Germany has developed quantum algorithms for weather-risk management by collaborating with the American computing company IBM. Iberdrola in Spain has developed quantum-inspired software via multiverse computing called Singularity, which is working to solve energy sector challenges like optimising power grids and improving generation planning. All of these projects combine classical simulation with quantum-ready algorithms. Onward to quantum While supercomputers tackle huge volumes of data at extraordinary speeds, quantum computers promise something more: solving problems that are practically impossible for even the biggest classical computers. Energy companies around the world are now investing in hybrid systems that combine quantum processors with traditional high-performance computing. This combination could unlock new ways to model complex chemical reactions, optimise power grids in real time, or predict maintenance needs across vast networks of pipelines and facilities. Quantum computing could also be a game changer for ambitions in nuclear fusion, the 'holy grail' of clean energy. Simulating the extreme physics inside a fusion reactor pushes classical computers to their limits - but quantum algorithms could help tackle challenges that stand in the way of viable commercial fusion. Across the sector, the implications are clear. From oil and gas majors to green energy pioneers, companies are investing heavily in digital twins, AI-powered reservoir modelling, and quantum pilot projects to refine exploration, cut emissions, and optimise trading operations. As global energy demand grows - and the push for net zero intensifies - supercomputers and quantum systems are fast becoming the unsung heroes behind the scenes. They're more than machines: they're the engines of a cleaner, smarter, and more sustainable energy future. Realities on the ground Last year Eni launched its new HPC6 system, which today ranks first in the world among supercomputers for industrial use, and the largest of any kind outside the US. It has a computing power of 606 PFlops peak (and 477 'sustained'), equal to over 600 million-billion complex mathematical operations per second. 'This is the sixth version of our HPC machine, and it has a huge amount of computational power,' says Pagani. 'The sheer power of the hardware is very important, and hardware can be provided by different suppliers according to the period, but what is even more important is having software that can perform as well as possible in this type of architecture.' He added: 'Therefore, in parallel with choosing the best hardware configurations that the supercomputing market can offer from time to time, we have been building highly specialised internal software skills to make best use of our computational power.' Eni is one of the few companies that has been building its machines rather than buying cloud computing services, which usually come from the United States. The code on which HPC6 runs started to be developed in the 1980s by Eni. Pagani says this is particularly important for Europe's energy security. 'My vision is that quantum computing could be very important for Europe, also to develop a new internet – a quantum internet,' he says. 'It's a new frontier where the EU can play a big role.' On datacenters, he says it's important for the EU to keep pushing for developing this type of technology and facilities in Europe to reinforce its data and computing sovereignty. Eni's Green Data Centre is a good example for the whole of Southern Europe of a highly efficient and continuously evolving asset whose capabilities can benefit not just Italy but our whole continent, he says. Green challenges However, there remains an irony that as much as supercomputing can help with the generation of energy, it is also a major energy user itself. So Europe's energy companies are also looking at ways to decrease the energy consumption of these facilities, even as the computing power increases. 'Our need was to have a significant internal capability in supercomputing, both hardware and software-wise', says Pagani. 'If you want to have this kind of capability, you first of all need to have your own 'house' for the supercomputer. We therefore decided in 2013 to realise our Green Data Centre, which has a very unique set of features. Since its construction, it's one of the most energy-efficient data centres in Europe and among the best for carbon footprint, because among its original requirements were high operational efficiency and minimisation of its environmental impact, placing therefore sustainability at the core of its design.' The challenge is that these supercomputers use a huge amount of power in a small space, which can be 10kw of power in just one square metre. As the supercomputers accelerate their computing power, innovation is necessary in the design to accommodate the space in an efficient way. Structures like Eni's Green Data Centre allow for refreshing the infrastructure as innovations are found to reduce energy use, even as the computing power increases. Partnerships This digital revolution is also fueling new partnerships. Many companies are now opening their computing power to startups, universities, and research labs to accelerate innovation across the entire energy ecosystem. The goal is to combine the expertise of data scientists, chemists, physicists, and engineers to tackle big challenges in renewable energy, energy storage, and emissions reduction. By opening advanced infrastructure to external researchers, the industry hopes to multiply breakthroughs - and prove that the energy transition will rely as much on bits and qubits as on barrels and wind turbines. But these companies are also asking policymakers for more support in making sure Europe is leading in this space, making the argument that it will improve Europe's energy security. [Edited By Brian Maguire | Euractiv's Advocacy Lab ]
Indian Express
5 days ago
- Automotive
- Indian Express
EaseMyTrip co-founder Prashant Pitti vows to cut Bengaluru traffic by 30% in one year
EaseMyTrip Co-founder Prashant Pitti Thursday announced a collaborative initiative aimed at reducing traffic congestion in Bengaluru by 25-30 per cent within a year. The plan, which brings together public institutions, private firms, and scientific minds on a common platform, was initiated following a series of high-level meetings over the past 10 days with key stakeholders, including Bengaluru traffic police, Bruhat Bengaluru Mahanagara Palike (BBMP), city police, Google India, IISc researchers, and mobility entrepreneurs. On social media platform X, Pitti described the collaboration as unprecedented, where for the first time all key public and private players will be working together. The initiative will rely heavily on data science, AI-driven simulations, and community involvement to enable real-time interventions and long-term systemic fixes. Among the immediate steps outlined, Pitti is exploring traffic simulation models with IISc and the traffic police's already established rerouting simulation tools that prioritise time over distance. Pitti has also approached Google, Uber, Ola, and Rapido to share anonymised data to strengthen predictive models. 'Once operational, these could anticipate and prevent gridlocks before they happen,' he said. Pitti has offered to take charge of an existing government grievance redressal app, expanding its scope beyond potholes to cover illegal parking, broken signals, waterlogging, and more. He said he was aiming to promote transparency by publicly showcasing complaints and resolution timelines. To prevent delays in road work due to unexpected rain, Pitti has also proposed a real-time weather predictor that will proactively address drainage-related disruptions. The approach signals a shift from infrastructure dependence to optimisation of existing systems using technology, intent, and data. 'If we continue to wait for infrastructure upgrades, we are merely playing catch-up,' Pitti said, emphasising that practical optimism and collaboration can solve what appears unfixable. Pitti has also launched a WhatsApp community to crowdsource feedback from daily commuters on trouble zones. He appealed to the public to contribute by identifying junctions with major congestion issues, helping the project pinpoint focus areas. This fresh announcement builds on an earlier offer made by Pitti on July 14, where he committed Rs 1 crore to fund a project aimed at identifying and solving choke points across Bengaluru using Google Maps' Road Management Insight, launched recently for city-level traffic analytics. The budget was proposed for hiring senior Machine Learning/Artificial Intelligence engineers, procuring Graphics Processing Unit resources and satellite imagery, funding Application Programming Interface calls, data storage, and creating detailed, time-based maps of chronic congestion zones.
Mint
7 days ago
- Mint
Do you really need an AI PC in 2025? -Explained-
AI is now baked into your browser, your apps, and now, your laptop. 'AI PC' has become the latest sticker on the box, with every major brand pushing models that come with dedicated AI hardware. But the real question is: are these next-gen machines actually useful, or is this just another spec bump wrapped in jargon? Let's break it down. An AI PC isn't just a laptop with smart features, it has hardware built specifically for AI workloads. The key ingredient is an NPU (Neural Processing Unit), a low-power chip designed to handle AI tasks locally, without relying on the cloud. This means things like real-time voice transcription, background blur, and AI summarisation happen on the device itself, faster and more securely. The current crop of AI PCs runs on Intel Core Ultra, AMD Ryzen AI 300, Qualcomm's Snapdragon X series, or Apple's M4 chip. These are the brains behind local Copilot access, smart camera effects in video calls, live captions, and even image generation tools. Because the timing finally makes sense. AI PCs are projected to account for nearly half of new laptop sales by the end of 2025. Demand for smarter features, improved battery life, and offline productivity has pushed manufacturers to go all in. You'll now find AI PCs across Windows (under the Copilot+ PC branding), macOS (with M4's improved ML cores), and even ARM-based machines. AI is becoming a standard. Performance boost: From quick photo clean-ups to running large language models locally, AI PCs make previously sluggish tasks seamless. Longer battery life: NPUs draw less power than CPUs or GPUs when running AI features, especially useful on the go. Better privacy: Since much of the AI processing happens on your device, there's less data being sent to external servers. Improved productivity: Copilot+ features like Recall, AI search across your files, and smart doc editing can genuinely save time, if you're the kind of user who'll use them. Future-proofing: AI workflows are only going to expand. Buying a capable machine now ensures you're ready for what's coming. You should consider an AI laptop if: You're a content creator, designer, or editor using apps that now come with AI baked in (like Adobe Firefly or DaVinci Resolve). You're in business or tech, juggling presentations, emails, meetings, and large datasets. You care about offline AI tools and don't want to depend on cloud services. You want a laptop that'll stay relevant for the next 3–5 years as AI capabilities grow. If your daily use revolves around Chrome, Excel, Netflix, and the occasional Word doc, you're not missing much. Also, AI laptops cost more upfront. You're paying a premium for power you might not use, yet. Budget-conscious buyers and students can comfortably opt for traditional laptops unless their workflow specifically benefits from on-device AI. If you're an early adopter or power user, AI PCs make sense right now. They're genuinely smarter, more efficient, and built for what's next. But if you're the average buyer, the smarter move might be to wait. Prices will fall, features will mature, and by 2026, most mainstream laptops will come with NPUs anyway.
Hamilton Spectator
22-07-2025
- Business
- Hamilton Spectator
Embedded LLM Launches First-of-its-Kind Monetisation Platform for AMD AI GPUs
SINGAPORE, July 21, 2025 (GLOBE NEWSWIRE) — Embedded LLM today announced the global launch of TokenVisor, its monetisation and management platform for Graphics Processing Units (GPUs). The platform was first unveiled alongside AMD at the Advancing AI 2025 conference (Santa Clara, California) held in June. As organizations race to build 'AI factories,' they face a universal challenge: how to translate massive hardware investment into measurable revenue or departmental value. Without turnkey tools for billing, usage tracking, and multi-tenant management, the path from capital expenditure to positive ROI is often slow and complex. TokenVisor addresses this challenge head-on, providing the first such commercialization layer specifically for the growing AMD AI ecosystem. 'TokenVisor brings powerful new capabilities to the AMD GPU neocloud ecosystem, helping providers efficiently manage and monetise LLM workloads,' said Mahesh Balasubramanian, Senior Director of Product Marketing, Data Center GPU Business, AMD. The platform's impact is already being recognized by major industry hardware leaders. 'TokenVisor flips the economics of AI infrastructure,' said Kumar Mitra, general manager and managing director of Lenovo in Greater Asia Pacific. 'By pairing Lenovo ThinkSystem servers with AMD Instinct GPUs and TokenVisor's turnkey monetisation layer, our customers are launching revenue-generating LLM services at unprecedented speed and scale, providing the financial guardrails and chargeback capabilities that CIOs and CFOs require to confidently greenlight AI investments at scale. It's the key to unlocking the full economic potential of the AI factory.' Engineered to simplify operations, TokenVisor enables GPU owners to: 'The spirit of open collaboration we saw at Advancing AI 2025 is what drives us. TokenVisor is the hypervisor for the AI Token era , born from that spirit and engineered with insights from the AMD neocloud community. It's the tool that unlocks the potential of decentralized GPU computing by making it commercially viable for everyone. This launch is a critical milestone, turning the vision of an open AI ecosystem into a reality for the entire AMD community.' Early Adoption Highlights Global Demand Early adopters consistently report that TokenVisor 'took the guesswork out of commercialization.' By providing built-in tools to benchmark capacity, control resource supply, and automate billing, providers can confidently model their business and launch revenue-generating services in a matter of days—a process that previously took months of custom development. The platform's comprehensive support for popular LLM models and responsive technical support are also cited as critical factors in this rapid deployment and return on investment (ROI). For media enquiries, please contact: Jiaqi Lim Head of PR & Marketing pr@ About Embedded LLM Embedded LLM PTE. LTD. creates innovative Large Language Model (LLM) platforms, empowering organisations with generative AI. Inspired by its mission to build essential AI infrastructure for the knowledge economy, the company delivers robust and secure solutions. A significant open-source contributor, notably enhancing vLLM for AMD ROCm, Embedded LLM also offers open source tools like its JamAI Base no-/low-code LLM orchestration platform. The company is committed to making LLM technology accessible and fostering innovation within the open ecosystem. About AMD For more than 55 years, AMD has driven innovation in high-performance computing, graphics, and visualization technologies. Hundreds of millions of consumers, Fortune 500 businesses, and leading scientific research facilities around the world rely on AMD technology to improve how they live, work, and play. AMD employees are focused on building leadership high-performance and adaptive products that push the boundaries of what is possible. For more information about how AMD is enabling today and inspiring tomorrow, visit . About Lenovo Lenovo is a US$69 billion revenue global technology powerhouse, ranked #248 in the Fortune Global 500, and serving millions of customers every day in 180 markets. Focused on a bold vision to deliver Smarter Technology for All, Lenovo has built on its success as the world's largest PC company with a full-stack portfolio of AI-enabled, AI-ready, and AI-optimized devices (PCs, workstations, smartphones, tablets), infrastructure (server, storage, edge, high performance computing and software defined infrastructure), software, solutions, and services. Lenovo's continued investment in world-changing innovation is building a more equitable, trustworthy, and smarter future for everyone, everywhere. Lenovo is listed on the Hong Kong stock exchange under Lenovo Group Limited (HKSE: 992) (ADR: LNVGY). To find out more visit . A photo accompanying this announcement is available at
Toronto Star
22-07-2025
- Business
- Toronto Star
Embedded LLM Launches First-of-its-Kind Monetisation Platform for AMD AI GPUs
The new platform signals a new phase of maturity for the AMD AI ecosystem, enabling providers to compete by rapidly deploying and billing for LLM services. Initially unveiled with AMD at the Advancing AI 2025 conference, the platform now launches globally to accelerate AI revenue for the entire neocloud ecosystem. SINGAPORE, July 21, 2025 (GLOBE NEWSWIRE) — Embedded LLM today announced the global launch of TokenVisor, its monetisation and management platform for Graphics Processing Units (GPUs). The platform was first unveiled alongside AMD at the Advancing AI 2025 conference (Santa Clara, California) held in June.
